CN1708133A - Method and apparatus for sub-pixel motion estimation which reduces bit precision - Google Patents

Method and apparatus for sub-pixel motion estimation which reduces bit precision Download PDF

Info

Publication number
CN1708133A
CN1708133A CNA2005100751088A CN200510075108A CN1708133A CN 1708133 A CN1708133 A CN 1708133A CN A2005100751088 A CNA2005100751088 A CN A2005100751088A CN 200510075108 A CN200510075108 A CN 200510075108A CN 1708133 A CN1708133 A CN 1708133A
Authority
CN
China
Prior art keywords
pixel
value
sub
difference
whole
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CNA2005100751088A
Other languages
Chinese (zh)
Other versions
CN100394799C (en
Inventor
朴灿植
李在宪
李男淑
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Samsung Electronics Co Ltd
Original Assignee
Samsung Electronics Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Samsung Electronics Co Ltd filed Critical Samsung Electronics Co Ltd
Publication of CN1708133A publication Critical patent/CN1708133A/en
Application granted granted Critical
Publication of CN100394799C publication Critical patent/CN100394799C/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
    • H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
    • H04N19/51—Motion estimation or motion compensation
    • H04N19/523—Motion estimation or motion compensation with sub-pixel accuracy
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
    • H04N19/184—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being bits, e.g. of the compressed video stream
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
    • H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
    • H04N19/51—Motion estimation or motion compensation

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Compression Or Coding Systems Of Tv Signals (AREA)

Abstract

A method and apparatus are provided for sub-pixel motion estimation which performs sub-pixel motion estimation in which the precision of a difference between a pixel value of a block to be searched and a pixel value of a block of a search area, i.e., the number of bits of the difference, is reduced. The method for sub-pixel motion estimation in inter-prediction includes (a) determining one integer pixel corresponding to an estimated block by performing integer-pixel motion estimation, (b) receiving a value of a sub-pixel that is adjacent to the determined integer pixel and a value of an integer pixel of a block to be searched, (c) reducing a difference between the value of the sub-pixel and the value of the integer pixel and reducing a bit precision of the difference, and (d) calculating a similarity using the difference whose bit precision is reduced.

Description

Reduce the method and apparatus that is used for sub-pixel motion estimation of position precision
The application requires to be submitted on June 11st, 2004 priority of the 10-2004-0042916 korean patent application of Korea S Department of Intellectual Property, and the present invention is disclosed in this for reference.
Technical field
Equipment consistent with the present invention and method relate to the estimation in the inter prediction, more specifically, relate to sub-pixel motion estimation, wherein, the precision of the difference between the pixel value of the pixel value of searched piece and the piece of the field of search, that is, the figure place of difference is reduced.
Background technology
When comparing with the video of encoding according to traditional coding method, the video of H.264 encoding according to video encoding standard is compressed with higher ratio, and has good quality.This is because whole pixel motion is estimated by sequentially carrying out, half-pix estimation and 1/4th pixel motions are estimated, the estimation in carrying out H.264 is so caused estimation more accurately.Yet, since the hardware of motion estimation unit account for whole encoder hardware 70%, so put on the very big load of encoder.In addition, because at half-pix and 1/4th pixels, that is, sub-pixel is carried out estimation, so compare with traditional estimation, this estimation needs more time.
Summary of the invention
The invention provides a kind of method and apparatus that is used for sub-pixel motion estimation, wherein, with the precision of the difference between the pixel value of the piece of the pixel value of searched piece and the field of search, that is, figure place is reduced in the sub-pixel motion estimation in the inter prediction.
According to an aspect of the present invention, provide a kind of method that is used for the sub-pixel motion estimation of inter prediction, this method comprises: (a) pass through to carry out whole pixel motion and estimate to determine a whole accordingly pixel of the piece with estimation; (b) receive the value of the sub-pixel adjacent and with the value of the whole pixel of searched piece with the whole pixel that should determine; (c) precision of difference between the value of the value of minimizing sub-pixel and whole pixel and reduction difference; (d) difference of using its precision to be lowered is calculated similitude.
Sub-pixel can be half-pix or 1/4th pixels, and half-pix inserts by whole pixel that will be adjacent with the whole pixel of determining and obtains, and 1/4th pixels can obtain by inserting the half-pix adjacent with the half-pix that obtains.
In (c), when poor between the value of the value of calculating sub-pixel and whole pixel, can delete by the high position of the predetermined number that will differ from and reduce poor figure place to reduce poor precision.
According to a further aspect in the invention, a kind of method that is used for the sub-pixel motion estimation of inter prediction is provided, and this method comprises: (a) receive the value of the sub-pixel adjacent with the whole pixel of estimating by whole pixel motion to determine and with the value of the whole pixel of searched piece; (b) carry out the degree that sub-pixel motion estimation is calculated the video quality reduction by the value of use sub-pixel and the value of whole pixel; (c) difference between the value of the value of calculating sub-pixel and whole pixel is carried out sub-pixel motion estimation and is also calculated the degree that video quality reduces to reduce poor precision; (d) when the degree of the reduction in (b) with the difference between the degree of the reduction (c) during less than the predetermined reference value, the search bit precision, poor according between the value of the value of position precision varitron pixel and whole pixel, and next video carried out sub-pixel motion estimation.
Can use Y-PSNR (PSNR) to calculate the degree that video quality reduces.
(d) the predetermined reference value in can and be defined in one of percentage position by the decibel unit representation.
According to a further aspect in the invention, the equipment that provides a kind of sub-pixel fortune merit that is used for inter prediction to estimate, this equipment comprises: the difference computing unit, be used to receive the value of the sub-pixel adjacent and the value of whole pixel that will searched piece with the whole pixel of determining, and poor between the value of the value of calculating sub-pixel and whole pixel; Precision reduces the unit, is used to reduce the precision of the difference of calculating; With the similitude computing unit, the difference that is used to use its precision to be lowered is calculated search block and with the similitude between the searched piece.
This equipment comprises that also the difference that its precision is lowered is converted into the converter unit of frequency domain.
This similitude computing unit can calculate absolute difference sum (SAD) at the difference that its precision is lowered, and perhaps calculates the summation (SATD) of absolute value of the difference of Ha Deman conversion at the result who is converted into frequency domain.
Description of drawings
By specifically describing exemplary embodiment of the present invention with reference to the accompanying drawings, above-mentioned and other aspects of the present invention will become apparent, wherein:
Fig. 1 is the diagrammatic sketch that is used to explain whole pixel motion estimation;
Fig. 2 is the diagrammatic sketch that is used to explain half-pix estimation and the estimation of 1/4th pixel motions;
Fig. 3 is the block diagram of motion estimation unit;
Fig. 4 has shown the exemplary configurations of PE array;
Fig. 5 has shown the concrete structure of PE;
Fig. 6 is the flow chart that the method that is used for sub-pixel motion estimation according to an exemplary embodiment of the present invention is shown;
Fig. 7 is the flow chart that the method that is used for sub-pixel motion estimation of another exemplary embodiment according to the present invention is shown; With
Fig. 8 is the block diagram that is used for the equipment of sub-pixel motion estimation according to an exemplary embodiment of the present invention.
Embodiment
Fig. 1 is the diagrammatic sketch that is used to explain whole pixel motion estimation.
With reference to previous frame, in order to estimate in present frame movement degree with the piece 110 of searched pre-sizing, promptly, the movement degree of macro block, the field of search 120 is selected from previous frame, be moved in the field of search 120 based on pixel with the piece of the big or small identical size of piece 110, and the pixel value of the piece in the pixel value of piece 110 and the field of search 120 is compared, therefore, the most similar piece of search and piece 110 in the frame formerly.In order to search for the piece the most similar to piece 110, for example, the piece that has from the minimum SAD of piece 110 is confirmed as the most similar piece, and is confirmed as estimating the whole pixel that obtains by whole pixel motion with the corresponding pixel of determining of piece.
After this, also, then 1/4th pixels adjacent with the half-pix of determining by the half-pix estimation are carried out estimation, to be used for more accurate movement estimation to carrying out estimation with estimate the adjacent half-pix of definite whole pixel by whole pixel motion.
Fig. 2 is the diagrammatic sketch that is used to explain half-pix estimation and the estimation of 1/4th pixel motions.
In Fig. 2, represent whole pixel by the square pixels of capitalization indication, represent half-pix by the circular pixel of lowercase indication, represent 1/4th pixels by the triangular pixel of numeral indication.
As described in reference Fig. 1,, then also the sub-pixel adjacent with this whole pixel carried out estimation if the whole pixel that whole pixel is carried out estimation and the most similar piece is determined.For example, if estimate that by whole pixel motion the whole pixel of determining is a pixel A 210, then to eight half-pixs adjacent, promptly with pixel A 210, pixel b, c, d, e, f, g, h and i carry out estimation, and the half-pix with piece of minimum SAD is determined.By in present frame with the pixel value of searched piece and formerly the absolute value read group total of the difference between the pixel value of the piece in the frame obtain SAD.Therefore, 1/4th pixels adjacent with the half-pix as above determined 220 are carried out estimation.
Same, owing to use whole pixel motion results estimated to carry out the half-pix estimation and 1/4th pixel motions are estimated, thus in the half-pix estimation with the value of the whole pixel of the piece in the value of the whole pixel of searched piece and the field of search is similar each other during 1/4th pixel motions are estimated.As a result, in that the difference between the value of the sub-pixel of the value of the whole pixel of searched piece and the piece in the field of search is little.Therefore, in half-pix estimation or the estimation of 1/4th pixel motions, there is no need to use this poor, but this difference can be used to calculate SAD to reduce the position precision that influences video quality a little.
Fig. 3 has shown the structure of motion estimation unit.
Motion estimation unit comprises pixel value memory cell 310, processing unit (PE) array 320 and determining unit 330.Pixel value memory cell 310 is stored the pixel value with the piece of the pixel value of searched piece and the field of search.PE array 320 receives pixel value and calculates similitude from pixel value memory cell 310.The summation (SATD) of the absolute value of the difference by calculating SAD or Ha Deman conversion obtains similitude.Determining unit 330 is determined and the corresponding pixel of motion vector according to the similitude of calculating in PE array 320, and the address of increase pixel value memory cell 310, so that pixel value memory cell 310 is sent to PE array 320 with next pixel.
Fig. 4 shows the exemplary configurations of PE array 320.
PE array 320 has bidimensional and has 8 * 8 processing unit PE among Fig. 4 for example.Each PE receives the pixel value of the piece of the pixel value of searched piece and the field of search and calculates SAD or the minimal processing unit of SATD.
Fig. 5 has shown the concrete structure of each PE.
Each PE calculates SAD or SATD in 4 * 4 module units.Each PE comprises four subtrator 510a to 510d, four absolute value calculation unit 520a to 520d, a plurality of adder unit 530a to 530d and Ha Deman converter unit 540.Each PE receives 4 * 4 pixel value with behavior unit.At first, from the pixel value S of 4 * 4 of the field of search first row 00, S 10, S 20, and S 30In deduct the pixel value C of current 4 * 4 first row 00, C 10, C 20, and C 30The result who subtracts each other is handled by Ha Deman converter unit 540, and is transformed frequency domain.Then, all additions of the absolute value that subtracts each other of frequency translation.At next clock period, from the pixel value S of the next line of 4 * 4 first row of the field of search 01, S 11, S 21, and S 31In deduct the pixel value C of the next line of current 4 * 4 first row 01, C 11, C 21, and C 31, and the result who subtracts each other is handled by Ha Deman converter unit 540.The absolute value that subtracts each other of the conversion all result of addition and addition is added to the addition result of the SATD that is used for current 4 * 4 previous row.Like this, in case the pixel value of next line is read out and above-mentioned processing is repeated, after four clocks, finish the calculating that is used for current 4 * 4 SATD so.
Can calculate SAD and not calculate SATD.In this case, Ha Deman converter unit 540 can not be included in the motion estimation unit.
Fig. 6 is the flow chart that illustrates according to the method that is used for sub-pixel motion estimation of exemplary embodiment of the present invention.
In the first operating procedure S610, estimate to determine a whole pixel by carrying out whole pixel motion.Having described whole pixel motion with reference to Fig. 1 estimates.In operating procedure 620, receive the value of the whole pixel of the value of the sub-pixel adjacent and search block with the whole pixel of estimating by whole pixel motion to determine.Each all is 8 for the value of this sub-pixel and the value that should put in order pixel, so the difference between them is formed by minimum nine.Therefore, in order to reduce the size of hardware, difference is treated to four or five in operating procedure S630.In other words, because whole pixel motion estimates to be performed before carrying out sub-pixel motion estimation, so the difference that is used between the input pixel value of sub-pixel motion estimation is little.Therefore, even be on duty when being treated to four or five, the video quality of coding also seldom is affected.In this exemplary embodiment, this difference is represented as four or five, but in some cases, it can be represented as still less or more position.Then, in operating procedure 640, the difference that its precision is lowered is used to calculate similitude, for example SAD or SATD.
Here, sub-pixel is half-pix or 1/4th pixels.Half-pix obtains by inserting the whole pixel adjacent with the whole pixel of estimating by whole pixel motion to determine./ 4th pixels obtain by inserting the half-pix adjacent with the half-pix of determining by the half-pix estimation.By removing poor several high positions, difference is treated to four or five to reduce poor figure place.
Fig. 7 is the flow chart that the method that is used for sub-pixel motion estimation of another exemplary embodiment according to the present invention is shown.
In the first operating procedure S710,, several frames calculate the degree that video quality reduces by being carried out traditional sub-pixel motion estimation.For example, the degree of reduction obtains by calculating PSNR.The traditional sub-pixel motion estimation meaning is to use poor between the pixel value.In other words, receive the value of the sub-pixel adjacent and the value of whole pixel that will searched piece, and calculate PSNR by the execution sub-pixel motion estimation with the whole pixel of estimating by whole pixel motion to determine, that is, and the degree of reduction.
In operating procedure S720, for identical frame, the precision of the difference between the value of sub-pixel, that is, figure place is lowered, and calculates the degree that video quality reduces by carrying out sub-pixel motion estimation.In operating procedure S730, the difference between the degree of the reduction of calculating in operating procedure S710 and S720 is than predetermined reference value hour, checks the figure place of difference between the value of sub-pixel.In operating procedure S740, for the frame of present frame back, the figure place of the difference between the value of sub-pixel is reduced to the figure place of inspection, and carries out sub-pixel motion estimation.For example, the PSNR that reference value has the 0.0A dB that determines in one of percentage decimal place is poor, and can be preferably 0.05dB.
Fig. 8 is the block diagram that is used for the equipment of sub-pixel motion estimation according to an exemplary embodiment of the present invention.
The equipment that is used for sub-pixel motion estimation comprises: difference computing unit 810, precision reduce unit 820, converter unit 830 and similitude computing unit 840.
Difference computing unit 810 receives the value of the sub-pixel adjacent with a whole pixel of estimating by whole pixel motion to determine and the value of whole pixel that will searched piece, and between the two poor of calculating.Precision reduces the figure place that unit 820 reduces the difference of calculating.For example, can reduce figure place to form four or five by deleting several high positions.Converter unit 830 is converted into frequency domain with the difference that its figure place is reduced.For example, can use the Ha Deman conversion to make such conversion.At this moment, when calculating SAD in order to calculate similitude, there is no need to use converter unit 830, but have only when calculating SATD, converter unit 830 needs.The difference that similitude computing unit 840 uses its figure place to be reduced is calculated similitude.
As the example of another equipment that is used for sub-pixel motion estimation, the equipment that is used for the estimation of sub-pixel fortune merit also can comprise determining unit 850.Determining unit 850 receives the value of the sub-pixel adjacent with the whole pixel of estimating by whole pixel motion to determine and only to a frame or a few frame value with the whole pixel of searched piece, when sub-pixel motion estimates that the figure place of the difference between the value of the value of the reduction degree of rear video quality and sub-pixel and whole pixel reduces that the difference between the sub-pixel motion estimation rear video quality reduction degree is less than predetermined reference value then, the figure place of the difference between the value of inspection sub-pixel, control precision reduces unit 820 figure place of the difference between the value of the value of sub-pixel and whole pixel is reduced to the figure place of minimizing, carries out sub-pixel motion estimation then.
The 0.0AdB that the predetermined reference value can preferably be determined two decimal places.
As mentioned above, according to the present invention, the size and the time quantum that are used for the required hardware of the estimation of video can be reduced greatly.
Therebetween, the method that is used for sub-pixel motion estimation can be implemented as computer program.The code and the section that constitute computer program can easily be explained by the computer programmer of this area.In addition, computer program is stored in the computer-readable medium, and is read and carry out by computer, therefore, realizes being used for the method for sub-pixel motion estimation.Computer readable recording medium storing program for performing comprises magnetic recording media, optical data storage device and carrier wave.
Though the present invention is specifically shown with reference to its exemplary embodiment and is illustrated, but will be understood by those skilled in the art that, under the situation that does not break away from the spirit and scope of the present invention that limit by following claim, can carry out the change of various forms and details to it.

Claims (19)

1, a kind of method that is used for the sub-pixel motion estimation of inter prediction, this method comprises:
(a) estimate to determine and estimative corresponding whole pixel by carrying out whole pixel motion;
(b) receive the value of the sub-pixel adjacent and with the value of the whole pixel of searched piece with the whole pixel that should determine;
(c) the position precision of difference between the value of the value of minimizing sub-pixel and whole pixel and reduction difference; With
(d) difference of using its precision to be lowered is calculated similitude.
2, the method for claim 1, wherein, sub-pixel is half-pix or 1/4th pixels, and this half-pix is determined by inserting the whole pixel adjacent with the whole pixel of determining, and this 1/4th pixel is determined by inserting the half-pix adjacent with this half-pix.
3, the method for claim 1, wherein in (c), when poor between the value of the value of calculating sub-pixel and whole pixel, delete by the high position of the predetermined number that will differ from and to reduce poor figure place to reduce poor position precision.
4, method as claimed in claim 3, wherein, in (c), the difference between the value of sub-pixel is treated to four or five.
5, a kind of method that is used for the sub-pixel motion estimation of inter prediction, this method comprises:
(a) receive the value of the sub-pixel adjacent and with the value of the whole pixel of searched piece with the whole pixel of estimating by whole pixel motion to determine;
(b) carry out the degree that sub-pixel motion estimation is calculated the video quality reduction by the value of use sub-pixel and the value of whole pixel;
(c) difference between the value of the value of calculating sub-pixel and whole pixel is carried out sub-pixel motion estimation and is also calculated the degree that video quality reduces to reduce poor position precision; With
(d) if the degree of the reduction (b) and (c) in the degree of reduction between difference less than the predetermined reference value, search bit precision then, poor according between the value of the value of position precision varitron pixel and whole pixel, and next video carried out sub-pixel motion estimation.
6, method as claimed in claim 5 wherein, at (c) with (d), uses Y-PSNR to calculate the degree that video quality reduces.
7, method as claimed in claim 6, wherein, (d) the predetermined reference value in is determined by the decibel unit representation and in one of percentage decimal place.
8, method as claimed in claim 5, wherein, sub-pixel is half-pix or 1/4th pixels, and this half-pix is determined by inserting the whole pixel adjacent with the whole pixel of determining, and this 1/4th pixel is determined by inserting the half-pix adjacent with this half-pix.
9, method as claimed in claim 5 wherein, in (c), when poor between the value of the value of calculating sub-pixel and whole pixel, reduces poor figure place by a high position of deleting poor predetermined number, to reduce poor position precision.
10, method as claimed in claim 9, wherein, in (c), the difference between the value of sub-pixel is treated to four or five.
11, a kind of equipment that is used for the sub-pixel motion estimation of inter prediction, this equipment comprises:
The difference computing unit is used to receive the value of the sub-pixel adjacent with the whole pixel of determining and the value of whole pixel that will searched piece, and poor between the value of the value of calculating sub-pixel and whole pixel;
Precision reduces the unit, is used to reduce the position precision of the difference of calculating; With
Similitude computing unit, the difference that is used to use its precision to be lowered calculate the similitude between search block and the searched piece.
12, equipment as claimed in claim 11 comprises that also the difference that its precision is lowered is converted into the converter unit of frequency domain.
13, equipment as claimed in claim 12, wherein, this converter unit is carried out the Ha Deman conversion.
14, equipment as claimed in claim 12, wherein, this similitude computing unit calculates the absolute difference sum at the difference that its precision is lowered, and perhaps calculates the summation of absolute value of the difference of Ha Deman conversion at the result who is converted into frequency domain.
15, equipment as claimed in claim 11, wherein, sub-pixel is half-pix or 1/4th pixels, and this half-pix is determined by inserting the whole pixel adjacent with the whole pixel of determining, and this 1/4th pixel is determined by inserting the adjacent half-pix of this half-pix.
16, equipment as claimed in claim 11, wherein, when poor between the value of the value of calculating sub-pixel and whole pixel, precision reduces the unit and deletes by the high position of the predetermined number that will differ from and reduce poor position precision.
17, equipment as claimed in claim 11, also comprise determining unit, be used to receive by whole pixel motion and estimate the value of the sub-pixel adjacent and the value of the whole pixel of the search block that is used for initial frame with the whole pixel of determining, when if the figure place of the difference between the value of the degree that sub-pixel motion estimation rear video quality reduces and the value of sub-pixel and whole pixel reduces difference between the degree that sub-pixel motion estimation rear video quality then reduces less than the predetermined reference value, then check the figure place of the difference between the value of sub-pixel, the figure place of the difference between the value of the value of sub-pixel and whole pixel is reduced to the figure place of minimizing, then next video is carried out sub-pixel motion estimation.
18, equipment as claimed in claim 17 wherein, uses Y-PSNR to calculate the degree that video quality reduces.
19, a kind of computer readable recording medium storing program for performing that has program recorded thereon thereon, this program is carried out the method for the sub-pixel motion estimation that is used for inter prediction, and this method comprises:
(a) estimate to determine and estimative corresponding whole pixel by carrying out whole pixel motion;
(b) receive the value of the sub-pixel adjacent and with the value of the whole pixel of searched piece with the whole pixel that should determine;
(c) the position precision of difference between the value of the value of minimizing sub-pixel and whole pixel and reduction difference; With
(d) difference of using its precision to be lowered is calculated similitude.
CNB2005100751088A 2004-06-11 2005-06-08 Method and apparatus for sub-pixel motion estimation which reduces bit precision Expired - Fee Related CN100394799C (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
KR1020040042916 2004-06-11
KR1020040042916A KR100580194B1 (en) 2004-06-11 2004-06-11 Sub pixel motion estimation method and apparatus reducing a bit precision

Publications (2)

Publication Number Publication Date
CN1708133A true CN1708133A (en) 2005-12-14
CN100394799C CN100394799C (en) 2008-06-11

Family

ID=35094406

Family Applications (1)

Application Number Title Priority Date Filing Date
CNB2005100751088A Expired - Fee Related CN100394799C (en) 2004-06-11 2005-06-08 Method and apparatus for sub-pixel motion estimation which reduces bit precision

Country Status (4)

Country Link
US (1) US20050276330A1 (en)
EP (1) EP1608180A1 (en)
KR (1) KR100580194B1 (en)
CN (1) CN100394799C (en)

Families Citing this family (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4536325B2 (en) * 2003-02-04 2010-09-01 ソニー株式会社 Image processing apparatus and method, recording medium, and program
US20060233258A1 (en) * 2005-04-15 2006-10-19 Microsoft Corporation Scalable motion estimation
US20070199011A1 (en) * 2006-02-17 2007-08-23 Sony Corporation System and method for high quality AVC encoding
US7912129B2 (en) * 2006-03-16 2011-03-22 Sony Corporation Uni-modal based fast half-pel and fast quarter-pel refinement for video encoding
US8077769B2 (en) * 2006-03-28 2011-12-13 Sony Corporation Method of reducing computations in transform and scaling processes in a digital video encoder using a threshold-based approach
US7929608B2 (en) * 2006-03-28 2011-04-19 Sony Corporation Method of reducing computations in intra-prediction and mode decision processes in a digital video encoder
US8494052B2 (en) * 2006-04-07 2013-07-23 Microsoft Corporation Dynamic selection of motion estimation search ranges and extended motion vector ranges
US8155195B2 (en) * 2006-04-07 2012-04-10 Microsoft Corporation Switching distortion metrics during motion estimation
US20070268964A1 (en) * 2006-05-22 2007-11-22 Microsoft Corporation Unit co-location-based motion estimation
US8023732B2 (en) * 2006-07-26 2011-09-20 Siemens Aktiengesellschaft Accelerated image registration by means of parallel processors
WO2008098194A2 (en) 2007-02-08 2008-08-14 Starent Networks Corporation System and method for handoffs between technologies
TW200845758A (en) * 2007-05-03 2008-11-16 Nat Univ Chung Cheng Fast intra coding algorithms for dynamic definition adjustment
KR100946775B1 (en) * 2008-01-28 2010-03-11 울산대학교 산학협력단 Fast search method for sub-pixel motion estimation in H.264
KR100994983B1 (en) * 2008-11-11 2010-11-18 한국전자통신연구원 Apparatus and method for estimation of high speed motion
JP5890794B2 (en) * 2013-02-28 2016-03-22 株式会社東芝 Image processing device

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2636622B2 (en) * 1992-03-13 1997-07-30 松下電器産業株式会社 Video signal encoding method and decoding method, and video signal encoding apparatus and decoding apparatus
KR100269129B1 (en) * 1997-11-21 2000-10-16 윤종용 Motion estimation method using subpixel-by-subpixel basis and apparatus therefor
KR100397055B1 (en) * 2000-07-21 2003-09-06 (주)씨앤에스 테크놀로지 Motion estimator architecture for low bit rate image communication
KR100727910B1 (en) * 2000-10-11 2007-06-13 삼성전자주식회사 Method and apparatus for motion estimation of hybrid type
JP2003348598A (en) * 2002-04-12 2003-12-05 Seiko Epson Corp Method and apparatus for memory efficient compressed domain video processing and for fast inverse motion compensation using factorization and integer approximation
CN1194553C (en) * 2002-04-18 2005-03-23 华为技术有限公司 Picture motion estimating method
KR100468726B1 (en) * 2002-04-18 2005-01-29 삼성전자주식회사 Apparatus and method for performing variable bit rate control in real time

Also Published As

Publication number Publication date
CN100394799C (en) 2008-06-11
EP1608180A1 (en) 2005-12-21
KR100580194B1 (en) 2006-05-16
KR20050117727A (en) 2005-12-15
US20050276330A1 (en) 2005-12-15

Similar Documents

Publication Publication Date Title
CN1708133A (en) Method and apparatus for sub-pixel motion estimation which reduces bit precision
CN1189037C (en) Motion estimation
US6377623B1 (en) High speed motion estimating method for real time moving image coding and apparatus therefor
US8218635B2 (en) Systolic-array based systems and methods for performing block matching in motion compensation
US11902563B2 (en) Encoding and decoding method and device, encoder side apparatus and decoder side apparatus
US20080101707A1 (en) Method for decomposing a video sequence frame
EP1775963A1 (en) Motion vector detecting device, and motion vector detecting method
CN1893665A (en) Error concealment method and apparatus
CN1956547A (en) Motion vector estimating device and motion vector estimating method
US7865026B2 (en) Data reuse method for blocking matching motion estimation
CN1325220A (en) Motion-vector coding method
CN110362964B (en) High-capacity reversible information hiding method based on multi-histogram modification
CN1627825A (en) Motion estimation method for motion picture encoding
US8243810B2 (en) Motion estimation approach for real-time embedded multimedia design
CN101044765A (en) Apparatus and method for image coding
CN1604650A (en) Method for hierarchical motion estimation
CN1852442A (en) Layering motion estimation method and super farge scale integrated circuit
CN1518365A (en) Noise estimating method and equipment, and method and equipment for coding video by it
CN1913632A (en) Motion picture encoding method and device
US8837595B2 (en) Motion estimation method
CN1201589C (en) Motion estimation
CN1787642A (en) Method for fast searching between frame of image unit module based on H.264
CN1625266A (en) Apparatus for calculating absolute difference value, and motion estimation apparatus and motion picture encoding apparatus
US7181075B2 (en) Multi-level encoding/decoding apparatus
CN1083217C (en) Start and end adresses generating circuit for motion compensation during moving picture compression coding

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
C17 Cessation of patent right
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20080611